Treatment implications 1 Source-separated urine

The collected findings indicate that at dilutions relevant for sanitation systems, components buffering against pH changes in urine seem to be

present at sufficiently high concentrations (Hellström et al., 1999). Provided a system has been in use long enough for urease-producing biofilm to develop, pH can be assumed to be 8.9 and above in any collection tank or container connected to a system with low flush. This is supported by most literature, reporting pH of 8.9-9.2 in sanitation systems of different degrees of dilution and TAN concentrations, in northern Europe (Chandran et al., 2009; Pradhan et al., 2009; Udert et al., 2003; Andersson & Jensen, 2002; Jönsson et al., 1997b; Kirchmann & Pettersson, 1995) as well as in Asia (Yang et al., 2003) and Africa (Paper III; (Mnkeni et al., 2006).

Dilution with flushwater will thus mainly affect the TAN concentrations. In Sweden, most urine-separating sanitation systems have some kind of flush system for the urine, but the amount used differs widely between different toilets and the user’s behaviour. In a Swedish eco-village, 16 to 290 mL water are used per flush, resulting in an average dilution of urine less than 3:1 with tap water. This is considered a low dilution as a result of committed users (Andersson & Jensen, 2002). Most other studies have assessed the dilution with flushwater to be 1:1 or 1:2. A 1:3 dilution with flushwater can be considered a high rate for a sanitation system and is only reported from one source, resulting in 1.7 g TAN L-1 (Udert et al., 2003). In general, TAN concentrations in urine collection tanks are higher than the 1.4-1.8 g L-1 measured in the present study with 1:3 dilution (Papers I & II). When urine is collected without any flushwater, TAN concentrations still vary due to diet and habits.

Volatile losses of TAN in urine-separating systems, mainly from the storage tank, can decrease the final urine nitrogen concentrations. In Sweden, such losses were estimated at max. 0.5% based on air exchange (Jönsson et al., 2000). At tropical temperatures, volatile losses during collection and storage can be expected to be higher. In the present studies losses were recorded at constant temperatures of 24 and 34 ºC but despite frequent sampling of treatment containers did not exceed 8% (Paper I).

Even if dilutions in the present studies related to each other regarding TAN (Paper I), assessment of degree of dilution seems to say little about total ammonia concentrations. Thus these should preferably be measured to make the right assumptions about pathogen inactivation rates. TAN concentrations can be estimated accurately in the field with quick sticks based on colorimetric methods. The ammonia concentration, which constitutes most of the nitrogen in urine, can then be used to provide farmers with information on nutrient content. The nitrogen content is important for the efficient use of urine, especially if mineral fertilisers are to be replaced (Tidåker, 2007).

Figure 12 shows that keeping urine undiluted can be crucial for reaching NH₃ concentrations sufficient for pathogen inactivation, i.e. between 40 and 60 mM at 24ºC and below. However, temperature itself is critical for ammonia sanitisation and at 14 ºC Ascaris eggs will be slowly inactivated despite fairly high NH₃ concentrations, i.e. as in undiluted urine. In countries with a temperate climate, e.g. Sweden, temperatures in large-scale underground storage are not likely to reach 20 ºC. However, in Sweden and Europe in general, human Ascaridiasis is rare and the risk related to fertilising with urine is probably very low.

0 50 100 150 200 250 0 5 10 15 20 25 30 35 40 Temperature (°C) NH 3 (mM ) 0 50 100 150 200 250 0 5 10 15 20 25 30 35 40 Temperature (°C) NH 3 (mM )

Figure 12. NH₃ formation in urine at 0-40 ºC based pH 9.0 and 6.8 g TAN L-1

(Jönsson et al., 2005) and dilutions 1:0 (black); 1:1 (grey) and 1:3 (shaded/white) thereof. Broken line shows NH₃ formed at the combination of 1 g N L-1

(94% assumed as TAN) and pH 8.8 (WHO, 2006). Circles give the NH3 concentrations in the present study at constant temperature

(dilution colour marked according to above) (Paper I) and shaded areas give the NH₃ for varying temperature at full sun location (lower), and incubator setting (upper) (Paper III).

Enhancing pathogen inactivation by increasing the urine temperature by exposure of containers to sun seems to be a good idea when collection containers are small (high surface to volume ratio high), storage capacity is limited or urine is used continuously, as in countries with several cropping

seasons per year. Placement of urine containers adjacent to a wall gave slightly higher urine temperature during the night and morning hours than exposure in the open and the thermal capacity of the wall may partly explain the lower peak temperatures (Paper III). If there are more than three hours of sun per day, as in the present study (Paper III), even higher temperatures can be achieved. Despite lower ambient temperatures, initial tests in Sweden have with enhancement reached water temperatures of 45 ºC in a 20 L container (Niwagaba, unpublished). Such further development of urine heating solar systems using with e.g. reflectors can be useful in regions with high prevalence of Ascaridiasis in combination with temperate climate, e.g. large parts of China and Latin America.

For safe use of urine after 6 months of storage at 20 ºC, as suggested by WHO (2006), it also needs to have an ammonia content of 60 mM to inactivate Ascaris eggs. If this concentration is achieved, e.g. by 4.2 g TAN L-1 and pH 8.8 or 2.9 g TAN L-1 and pH 9.0, the storage time can be shortened to 1 month at a temperature of 34 ºC. When urine temperature varies above 20 ºC, the mean temperature could be used to relate to recommendations based on constant temperatures and would then most likely give a higher pathogen inactivation than the constant temperature (Paper III). At temperatures below 20 ºC, no assurance can be given that urine is safe regarding content of viruses (WHO, 2006) or Ascaris eggs (Paper II). However, at NH₃ concentrations of 60 mM (in the present study achieved in undiluted urine at 14 ºC) restricted use to crops not intended for human consumption can be accepted after only 1 month of storage, even at temperatures as low as 4 ºC, as the two major zoonotic risk organisms in urine, Salmonella and Cryptosporidium, will be inactivated (Höglund & Stenström, 1999; Jenkins et al., 1998); Paper I).

Urine at container bottoms has higher concentrations of organisms (Höglund et al., 2000) especially Ascaris and other parasitic eggs and cysts, which sediment easily (Panicker and Krishnamoorthi, 1981). To avoid cooling effects from the surface, as observed in the present study (Paper III), the cans can be put on a pallet or on some insulating material, since even small increases in temperature can contribute to Ascaris inactivation. In general, avoiding fertilising with urine bottom sludge could be a means to avoid risks from Ascaris if sanitisation is not achieved.

To generalise, having sanitation systems that keep the urine as concentrated as possible, i.e. no or low flush systems, is the simplest way to enhance pathogen inactivation in source-separated urine. Another positive effect from collecting urine undiluted is that systems without flushwater seem less prone to clogging (Udert et al., 2003b).

7.4.2 Faecal urea treatment

Adding urea to faecal material gives an alkaline pH, important for NH₃ formation, as well as increasing the total ammonia concentration (Paper IV). The collected findings indicate that pH may increase in proportion to the rate of urea addition until a breakpoint is reached where further additions will contribute little to further pH increase. The composition of the untreated material seems to determine the pH reached and its stability. Similar material dependency is observed when ammonia solution is used, but the resulting pH at breakpoint is higher than for urea (Ottoson et al., 2008a; Mendez et al., 2002; Allievi et al., 1994). To generalise, urea will give a pH above 9 and ammonia solution a pH above 10, provided the breakpoint is reached.

Lower TS through dilution would probably result in higher pH, although this conclusion could not be drawn when comparing urea added to different materials. Until knowledge gaps are filled, urea is recommended to be added at rates higher than 0.5% to faecal matter with TS as high as in the present study, 17-20%.

Two approaches can be envisaged with urea and ammonia treatment: to strive to keep additions and costs at a minimum level or to adjust the final nitrogen content to the fertiliser application. For the former it may be wise to monitor pH, since it may decrease to a level critical for ammonia sanitisation if urea or ammonia addition is low. With the latter it is important to use TAN concentrations that are applicable to common operational practices and equipment at the farm. With current fertiliser equipment the concentration should not be higher than about 10 g TAN per litre, corresponding to urea addition of 2% w/w or 1% ammonia from solution.

Dry matter may affect the mixing of the ammonia amendment into the material. A dry matter content above 20% was found to be critical for intermixing, and thus Ascaris egg inactivation, when cesspools were disinfected with liquid ammonia (1.5 to 6%) (Chefranova et al., 1984). During collection in a dry system, unless flushwater, urine or accidental water enters the collection bins, the moisture content will decrease from the initial 10-23% (Lentner & Geigy, 1981). In the present studies the collected faeces contained 25.7 to 27.0% TS, which was standardised to 17-20% before use. Other faeces of the same origin were reported to have a TS of 18-23.5% (Vinnerås et al., 2003a; Andersson & Jensen, 2002; Vinnerås & Jönsson, 2002), despite including toilet paper, in contrast to the present study. Urea can be assumed at the large scale to encounter similar problems with mixing at high TS, but water can be added to enhance mixing. Urea

distribution and decomposition have been shown to increase with increasing moisture content (Vinnerås et al., 2009).

However, at equal urea addition made on wet basis, a higher TS will result in higher TAN concentration in the solute when the urea has decomposed. A higher solute concentration may account for lower pH in material high in TS. However, depending on texture, a material with high TS may be easier to mix with urea. Mature compost with 75-82% TS supplemented with 6.5% wet-weight urea with the aim of enhancing the fertiliser value also resulted in sanitisation of the material (Adamtey et al., 2009).

It seems that urea is not likely to be 100% degraded (Paper IV; Sylwan, 2010; Agostini, unpublished). If sanitisation estimates are based on the present treatment additions of urea (Table 9), such non-degraded urea is accounted for within the recommendation. Even if not all nitrogen added by urea can be accounted for in the sanitisation, the nutrient value has to be based on all nitrogen added.

Increasing the pH with ash or lime or other alkaline agent can result in substantial NH₃ formation in faecal matter. The high pH of 12.8 in the present study cannot be considered representative due to the sieving, whereas the pH of 10.5 reached with unsieved ash (1 L kg-1

faeces) is more likely (Tesfaye, 2009; Moe & Izurieta, 2003).However, when ash is added continuously during collection, ammonia nitrogen is at high risk of being lost and the effect from increased pH will probably be lower that that seen in the present ash studies, where the containers were closed upon the addition of ash. Adjustment of the moisture content in the ash treatments, beneficial for a homogeneous distribution of hydroxide ions, may also explain the higher inactivation of Ascaris eggs in relation to pH compared with other studies (Sanguinetti et al., 2007).

The high pH from the ash probably inhibited the hydrolysis of the urea (Kabdasli et al., 2006), explaining very low ammonia recovery (Paper II). Therefore faecal material collected with ash amendments might not be suitable for further treatment with urea. Fidjeland (2010) added ash to faeces after 4% urea had first decomposed but concluded that water added to enhance distribution of hydroxide ions resulted in the final NH₃ concentration being the same as in faeces with only urea added.

The present studies showed ammonia sanitisation to be highly efficient at higher temperatures. At 34 ºC the S. Typhimurium 28B phage, which can be considered a conservative indicator according to enteric viruses, showed that the treatment time could be reduced from the recommended 1 year (WHO) to 2 and 1 month, with 1 and 2% urea treatment, respectively. At

such treatments Salmonella spp. and E. coli will be eliminated (i.e. a 12 log₁₀ reduction). However, at 24 ºC markedly slower inactivation took place despite urea treatment, mainly for the 28B phage.

0 40 80 120 160 200 240 280 320 360 400 440 480 0 10 20 30 40 Temperature (°C) NH 3 (m M )

Figure 13. NH3 concentrations reached at respective temperature in the present faecal

treatments with 2 (■), 1 (■) and 0.5% (□) urea, in relation to assessed concentrations from urea additions assuming 80% degradation, pH 8.9 and intrinsic TAN default values from Jönsson et al. (2005). Untreated faces assumed to hold pH 8.

Addition of 1% urea to faeces at temperatures from 14 to 34 ºC is sufficient to produce a safe fertiliser for unrestricted use (6 log₁₀ pathogen reduction) within 2 months of treatment as regards Salmonella spp., which would most probably also ensure inactivation of other enteric bacterial pathogens such as E. coli O157:H7. Addition of 2% urea at 24 and 34 ºC produces a safe fertiliser for unrestricted use within 8 months and 1 month, respectively, as regards the Salm. Typhimurium bacteriophage 28B, which was more persistent than Ascaris eggs at those temperatures (Table 3 and Table 8). At 14 ºC, neither Ascaris eggs nor any bacteriophages were studied in faeces but treatment time of Salmonella spp. was markedly shortened, from 2 months to less than 2 weeks, by increasing urea addition from 1 to 2%, respectively. The organism inactivation in urine indicates the importance of having as high ammonia concentration as possible at temperatures of 24 º and below.

A combination of pH and ammonia was highly efficient in inactivating all organisms studied at the higher temperatures. Alkaline amendment after urea is degraded or in combination with ammonia solution can be a means to achieve faster pathogen reduction at the lower concentrations without increasing the nitrogen content in the excreta fertiliser.

7.4.3 Sanitising other biomaterial

The results presented here regarding the inactivation of the pathogens Salmonella spp., E. coli O157:H7 and Ascaris eggs were related to NH₃ and temperature rather than material. Since nitrogen in human excreta and manure is partly in the form of ammonia or urea, a shift in waste management can enhance both nitrogen conservation and pathogen inactivation. Incompletely sanitised manure has been identified as an important factor for Salmonella spp. occurrence in animal herds (Cardinale et al., 2004; Veling et al., 2002) and manure treatment can prevent recirculation within herds and ensure food security. When the SEPA proposal on regulation of sewage product quality comes into action, sewage sludge treatments plants will have to include a sanitisation step in their treatment. Due to the high costs of heating and to increasing interest in the use of biogas, alternatives to heat treatments are highly relevant (Sylwan, 2010). In sparsely populated areas, separate collection of toilet waste in closed tanks may be the only option to protect sensitive water recipients. Such material, high in TAN, is very suitable to be sanitised with ammonia. Pathogen inactivation studies in manure from swine (Vinnerås, 2007) and cattle (Ottoson et al., 2008a; Diez-Gonzalez et al., 2000), as well as sewage sludge (Sylwan, 2010; Pecson & Nelson, 2005) and compost (Adamtey et al., 2009), all treated with urea or aqueous ammonia, indicate the applicability of ammonia sanitisation to biowaste in general.